| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Division of Pulmonary and Critical Care Medicine, Cedars-Sinai Medical Center, and David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, California
Correspondence and requests for reprints should be addressed to Zab Mosenifar, M.D., Department of Medicine, Division of Pulmonary/Critical Care Medicine, Cedars-Sinai Medical Center, 8700 Beverly Blvd., Room 6732, Los Angeles, CA 90049. E-mail: mosenifarz{at}cshs.org
ABSTRACT
Inclusion of underrepresented groups in clinical trials is important for several reasons. Age, sex, race, genetic factors, concomitant use of other medications, and comorbid conditions all may play pivotal roles in response to a drug or intervention. Despite the legislation for broader inclusion of underrepresented groups in clinical trials (via the National Institutes of Health [NIH] Revitalization Act of 1993), underrepresentation of particular populations, particularly minorities, continues to be a problem. Studies of predictors of clinical trial enrollment suggest that most people participate in clinical research to find relief from a disease, not for financial remuneration. Yet, men and whites are more likely to enroll in studies and some data indicate that certain patient populations are preferentially (albeit sometimes inadvertently) chosen for study enrollment. This tendency toward inclusion stems from human nature—the natural tendency for an investigator to relate to a particular investigative topic due to a special connection based on a cultural, socioeconomic, age, ethnicity, or gender level. This article reviews the most common population issues for clinical studies: age, gender, race, socioeconomic status, comorbidities, and disease severity, with examples of each from published studies. Recommendations are also offered to overcome these barriers.
Key Words: ethics • clinical trial • minority group • bias • gender gap
In 1993, Congress passed the National Institutes of Health (NIH) Revitalization Act, which required the Director of the NIH, along with the Director of the Office of Research on Women's Health and the Director of the Office of Research on Minority Health, to establish guidelines on inclusion of women and minorities as subjects in clinical research supported by the NIH. In 1994, the Federal Register announced the establishment of those guidelines (1). Because women and minorities had hitherto been under-represented, their inclusion in federally funded clinical research was now mandated.
Inclusion of minorities (or underrepresented groups) in clinical trials is important for several reasons. Age, sex, race, genetic factors, concomitant use of other medications, and comorbid conditions will have an impact on the nature of the response to various interventions. Nonetheless, despite the legislative power of different agencies, underrepresentation of particular populations, particularly minorities, continues to be a problem.
PREDICTORS OF CLINICAL STUDY ENROLLMENT
In 1999, 675,000 people participated in clinical research; they were chosen from 6.5 million individuals who were originally contacted. Of the 675,000 participants, 1,050 (0.16%) completed a survey regarding the reasons why they chose to participate in clinical studies. A total of 60% participated in research to find relief from a disease, 30% to advance science, and 10% to receive the stipend. Therefore, it is interesting to note that in this very sample of responders financial incentives played only a minor role in reasons to participate in clinical research. The concerns about research included fear of side effects, feeling like a "guinea pig," and the lack of follow-up after the clinical study is terminated (2). Citing only a 5% participation rate in clinical trials among patients with newly diagnosed cancer, Du and coworkers reviewed the medical records of 427 patients with lung cancer to determine predictors of clinical trial enrollment (3). They found that enrollment was significantly associated with race (P < 0.001), gender (P = 0.048), age (P = 0.005), and insurance type (P = 0.024). After multivariable adjustment, race and gender remained significant. The odds ratio for enrollment for African Americans was 0.485 (95% confidence interval [CI], 0.24–0.96) and for men it was 1.81 (95% CI, 1.03–3.1) (3). Thus, men were almost twice as likely to enroll as women, and African Americans were less than half as likely as whites to enroll. Indeed, 12 years after the announcement in the Federal Register, these disparities continue. The most common population issues for clinical studies are age, gender, race, socioeconomic status, comorbidities, and disease severity. Examples of each are given below.
POPULATION ISSUES
Age
Gurwitz and colleagues conducted a systematic review of clinical studies of drug therapies for treatment of acute myocardial infarction (MI), published between 1960 and 1991 (4). Of the 214 trials (involving 150,920 study subjects) they analyzed, 60% excluded persons older than 75 years of age. Of note, studies published after 1980 were more likely to have this age exclusion than studies published before 1980. The studies with age-based exclusions also had fewer female enrollees (18% versus 23%, P = 0.0002). As Gurwitz and coworkers note, the elderly experience the most morbidity and mortality from acute MI, so their exclusion from clinical trials of acute MI treatments clearly limits the ability to generalize the study results (4). Age is an important issue for many critical illnesses, especially respiratory illnesses, so their exclusion for trials in critically ill patients is an important problem. Researchers in general tend to impose an upper limit on age in their studies because the higher incidence of co-morbidities in the elderly might lead to higher levels of confounding findings.
Gender
The U.S. Food and Drug Administration (FDA) analyzed gender distributions among 493,000 subjects enrolled in 2,581 clinical trials from 1995 to 1999. Evelyn and coworkers found the proportion of women fluctuated each year, ranging from 42% in 1995 to 55% in 1998, with no directional trend (5). Their participation was not consistent. When the data were considered by specific disciplines, women and men were essentially equally represented in clinical trials in endocrinology, pulmonary medicine, and special pathogens. Women comprised less than half of the participants in four disciplines: oncology, medical imaging, cardio-renal, and antiretroviral studies.
As mentioned at the beginning of this article, the early 1990s brought legislation to mandate inclusion of women and minorities in clinical trials. However, as early as the 1960s and 1970s, protectionist legislation to exclude women was enacted because of the risk of fetal injury. In the 1980s, inequities in AIDS clinical trials were noted. The most compelling reason for excluding a gender from a clinical trial is if there is a high concentration of a disease in one gender. Otherwise, if no gender difference is anticipated, there is no compelling reason to exclude a gender from a clinical study (6).
A report from the National Academy of Sciences (NAS) on women and health research acknowledges an unconscious bias toward inclusiveness (i.e., male investigator tends to include people who are more like the investigator; similarly, women may feel uncomfortable with male researchers). Societies are stratified by gender and race, and different groups within societies also stratify along various lines. The report notes the difficulties (if not near impossibility) of disentangling the research process from the world within which it is conducted. Thus, one way to avoid the subtle bias of inclusion in a white male–dominated scientific research community is to encourage more women and minorities to become researchers (6).
The NAS report also discusses a law case suggesting that if a drug was found to harm women and yet women were excluded from the study, the study sponsor might still be found liable. Since 1990, significant effort has gone toward opening eligibility for clinical studies to pregnant women if they have full information about the risks to the offspring. They may be excluded if there is substantial evidence of harm to the offspring or there is no benefit to the study subject. However, most investigational review boards feel uncomfortable with the issue of clinical trial research in pregnant women (6).
Race
Race issues in clinical research are even more controversial. As discussed by King in an editorial, the barriers to race disparities exist at multiple levels: patient, care providers, the institution, and the community. King also comments on the many factors that affect physician–patient relationships that are often not consciously perceived. For example, many African Americans view signing the informed consent as relinquishing their autonomy and providing legal protection to physicians or the research institution (7, 8). Such prejudicial attitudes create mistrust among minorities of both physicians and the overall healthcare system. Such mistrust leads to reluctance to comply with medical regimens, and poor compliance leads researchers to avoid recruiting marginalized individuals and a propensity toward finding an ideal study subject. For example, in an HIV-related clinical trial, a "good participant" from a clinical scientist's narrow viewpoint is a white male, a risk taker, who is college educated, employed, housed, middle-class, and has insurance (7). Hence, the unspoken and unrecognized attitudes we as human beings carry with us have profound effects on clinical research (8).
Ethnicity can also affect study eligibility and interpretation of results. In pulmonary medicine, for example, the predictive equations for lung functions are different between African Americans and Asian Americans. Most studies do not consider these differences when analyzing efficacy data, especially if the results are expressed as absolute numbers. Similarly, if the selection criteria in a pulmonary clinical trial excludes subjects with forced expiratory volumes in 1 second of less than 70%, that might be a normal value for one population (e.g., ethnicity, geographic region) and subnormal for others. Because most laboratories do not use different predictive equations for minorities, this exclusion would lead to inadvertent selection bias and eventually suboptimal conclusions.
Socioeconomic Status
Socioeconomic status also creates barriers to research participation, for both social and logistical reasons. For example, homelessness, lack of transportation, and lack of fluency in English can preclude reliable participation in a clinical trial in the U.S. In a study of Asians, 79 elderly Asian immigrants from Taiwan, China, or Hong Kong were compared with 58 Asian American older adults regarding their responses to hypothetical clinical research situations. The study showed that the immigrant group was more likely to be influenced by a request from a son/daughter, landlord, physician, or advertisement (P < 0.001) and by a monetary incentive (P = 0.05) than the Asian Americans to participate in a clinical study. In other words, the power of persuasion was markedly different in people with the same ethnicity but different cultural influences. The authors conclude that "acculturation or assimilation into American society may build resistance to pressure to participate in research. Our findings also suggest that elderly Asian immigrants may need additional protections to achieve truly informed consent" (9).
Susceptibility to persuasion is an important issue in clinical trial recruitment, especially as cultures differ in familism (i.e., family-centered decision making process), their reverence for authority (including views of physicians as authoritative figures), and adoption of a "wait and see" attitude (i.e., "I don't want to be a guinea pig. If it works in other people, then I'll try it.") (9).
Biology
The potential for racial differences in drug response was discussed nearly 20 years ago (10). Svensson considered both the under-representation and disproportionate use of minorities in clinical studies, the latter arising from the proximity of most university hospitals to inner city, urban areas. Svensson's survey found that investigators do not adequately take into account racial differences as a potential source of variability. Thus, if a study includes a disproportionately large number of minorities and a difference in outcomes is observed, the results may be skewed in favor of a drug that may not occur when given to a broader population. Svensson concludes that "This under-representation in clinical trials suggests that insufficient data exist to accurately assess the safety and efficacy of many new drugs in American blacks" (10).
An important example of biological effects with ethnicities was observed in the Salmeterol Multicenter Asthma Research Trial (SMART), a 28-week, randomized, double-blind, placebo-controlled study comparing the safety of salmeterol to placebo in addition to usual asthma care in 26,355 subjects. The study subjects were 64% female, 36% male, and with the following ethnic profile: white, 71%; African American, 18%; Hispanic, 8%; other, 3%. In fact, the ethnic mix was fairly representative of U.S. society. The study outcomes showed no significant difference in the number of respiratory deaths or intubations between the two treatment groups: 50 versus 36. However, in African American subjects, the difference in this outcome between the two treatment groups was significant: 20 versus 5 (11).
Given these results, how should the pulmonologist treat the next African American patient with asthma? Do these data warrant a black box warning? Should a new clinical trial be conducted? Should all clinical trials examine the impact of race and ethnicity? If significant differences appear, a new trial study should investigate the effect in that subgroup.
Comorbid Diseases
The presence of comorbid diseases is a rate-limiting factor in study design. For example, African Americans have the highest mortality rates and poorest survival from cancer compared with other ethnic groups. A recent study followed 235 consecutive African-American patients with cancer who were considered for recruitment onto cancer treatment clinical trials at Howard University Cancer Center between January 1, 2001, and December 31, 2002. Of the 235 subjects, eligibility rate was 8.5% (n = 20); of these 20, enrollment was only 60% (n = 12). Comorbidities excluded 17% of the 235 patients and advanced disease made an additional 10% ineligible. In addition, the presence of HIV or anemia rendered an additional 37% of the patients ineligible (12). To overcome these barriers, investigators must cast the net wide to find as many eligible minority patients as possible. Some argue that clinical studies should have a narrowly defined population to increase the statistical power of the results. Others argue that recruitment should be as inclusive as possible, but the findings may be difficult to interpret. Herein lays another facet of clinical equipoise: balancing the need for generalizability of results with the need for statistical rigor. Most trials, however, make an attempt to exclude the major comorbidities such as cancer in sepsis trials, pulmonary hypertension in emphysema trials, and the use of steroids in reversibility trials.
RECOMMENDATIONS
DISCUSSION HIGHLIGHTS
Strategies for Recruiting Minorities into Clinical Trials
DR. DAVID COULTAS (The University of Texas Health Center at Tyler): My experience in the minority communities has been that you have to be a visible part of that community and willing to show a long-term relationship and commitment.
DR. MOSENIFAR: It's an issue of the long-term lack of trust, the trust that's been lost for many, many years. You really have to find an exclusive disease, so there would be no misunderstanding in why blacks were being recruited specifically to participate. It's very difficult to build that trust overnight.
Minority Issues with Respiratory Clinical Trials
DR. MOSENIFAR: Besides genetic issues (such as polymorphism) or the pattern of smoking in African Americans, I don't know of any particular issue in recruiting for respiratory clinical trials, or conducing clinical trials for respiratory disorders.
DR. GERARD J. CRINER (Temple Lung Center, Philadelphia, PA): For COPD, not thinking about race and recruitment has hurt the outcomes. The perception is that blacks are not as afflicted by COPD as whites, but they lead in the number of people with ED visits and hospital admissions (based on racial demographics) in COPD. So, it's important to include different races, for several reasons: genetics/drug metabolism differences, social environment, or access to care.
DR. COULTAS: We also have to consider some of the social anthropology. For example, Native Americans would not accept randomization to a placebo—the idea of a placebo, or something that is designed not to work, is unacceptable to them.
DR. MOSENIFAR: In one sense, though, "placebo" is the wrong term. We may be hurting ourselves by using the word "placebo." If they're getting usual care, that's the control. Control is not getting nothing; it's getting their conventional therapies. When we say "placebo," people have this image, wrongly, that they are getting only an inert treatment, but in fact they are getting their regular treatment.
DR. COULTAS: That goes back to culture and health literacy, and there is a whole field of literature outside of medicine—medical anthropology.
Role of Government and Minority Issues in Clinical Trials
MR. JAMES K. HAVEMAN (Haveman Group, Grand Haven, MI): As a commissioner, I think we have been excluded purposely from a lot of discussion on clinical trials, and it has a huge impact on what we're trying to do. For asthma within the African-American community, and with certain populations, we haven't had the ability to bring in people from the clinical trials to assist us, but we would welcome some partnerships.
My fear is that if those who are doing clinical trials don't more fully disclose who is participating in the study, how the money is going to the university, how the physicians gain from the study, who is writing the report, we're going to see legislatures step in and pass statewide legislation, because the financial impact is too great for them not to look at it. It's something that people should take seriously.
I think you could do a better job by using state physicians and people who partner with universities to be included in the design phase. I think that would serve you well. It would bring some friends to the table, if we were involved more and earlier in the design of some clinical studies in routine care.
FOOTNOTES
This article was funded by Boehringer Ingelheim and by Pfizer, Inc.
Conflict of Interest Statement: Z.M. has received $4,500 in 2004, 2005, and 2006 from Boehringer-Ingelheim and $4,000 from Pfizer in 2006 for speaking at conferences organized by these companies. He is an investigator in the uplift trial. The amount of this grant is $11,000 over a period of 4 years. Z.M. received no salary or consulting fees, however. D.C. received $12,000 in 2005 for speaking at conferences sponsored by Pfizer, and $2,000 for speaking at the conference on research ethics sponsored by Boehringer-Ingelheim and Pfizer, which formed the basis for his manuscript. G.J.C. has received research grants from Boehringer-Ingelheim, GlaxoSmithKline, Novartis, Schering-Plough, Emphasys, and Aeris. He has also participated on Advisory CTES of Otsuka, Ortho Biotec, and Schering-Plough. None of these activities are related to the subject matter of this manuscript. J.K.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
(Received in original form January 3, 2007; accepted in final form February 14, 2007)
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
